This is the mail archive of the
binutils@sourceware.org
mailing list for the binutils project.
Re: [Patch] Add --identify option to dlltool
- From: Kai Tietz <Kai dot Tietz at onevision dot com>
- To: Charles Wilson <cygwin at cwilson dot fastmail dot fm>
- Cc: binutils at sourceware dot org
- Date: Tue, 11 Nov 2008 10:04:16 +0100
- Subject: Re: [Patch] Add --identify option to dlltool
> I guess there are two choices:
>
> 1) ignore it. Only return "the first" .idata$7 entry.
>
> 2) return all entries (one per line?). In this case, a client (such as
> libtool) could detect a multi-line result and flag an error, or
> otherwise handle that exceptional case.
>
> The downside of 2) is that in the VAST majority of cases, there is only
> one archive member that has the non-empty .idata$7 section -- and it is
> first. So, --identify usually quits after parsing only the very first
> archive member; it hunts only until it finds the (one, e.g. first, e.g
> only) answer. Option (2) would require parsing *every* archive member
> from beginning to end. That's a pretty severe pessimism to handle a case
> that is rather pathological in nature.
IMHO case 1) is that one to choose. Maybe it would be even better to check
if there is just one .idata$7 section present, and if not to reject it at
all. The example I made is for *nix an absolute none-issue, I agree. For
coff targets it is possible but odd and shouldn't be supported at all, but
should be warned.
Cheers
Kai
| (\_/) This is Bunny. Copy and paste Bunny
| (='.'=) into your signature to help him gain
| (")_(") world domination.